PROJECT SUMMARY Autism, a neurodevelopmental disorder characterized by impairment in communication, social impairment and repetitive and stereotyped behaviors, is a life-long condition with an undetermined etiology and currently is not curable. While autism has a heterogeneous and complex genetic underpinning there is evidence that environmental factors also play a role. Despite the variety of factors that can lead to autism, the phenotype is remarkable uniform, raising the possibility of a ?final common pathway? in this disorder. Functional MRI and electrophysiology studies suggest this ?final common pathway? may be through aberrant neural connectivity during development. In this proposal we wish to evaluate previously recorded EEGs obtained from a large cohort of children with autism, developmental delay and neurotypically developing children extensively evaluated at the NIMH, to determine if there are neural networks characteristics that distinguish children with classic autism from typically developing children and children with developmental disabilities without autism. Ascertaining such changes may provide insight into the pathophysiological mechanisms responsible for the symptoms in autism. In specific aim 1 we will determine whether children with classic autism have neural networks which distinguish them from typically developing children and children with neurodevelopmental disorders but without autism. We will examine awake and sleep recordings for coherence and develop functional connectivity maps using Pearson correlations and partial correlations in all three groups. In specific aim 2 we will determine whether coherence and connectivity maps change over time in children with autism and whether such changes correlate with outcome and in specific aim 3 we will determine whether epileptiform activity is more common in the EEGs of the children with autism than in typically developing children and children with neurodevelopmental disorders but without autism and whether such activity is related to outcome. By taking advantage of this rich data set we wish to better characterize neural connectivity in autism using powerful electrophysiological techniques. Understanding how the brain of a child with autism is ?wired? will play an important role in developing therapeutic interventions.